基于分子信标微流控芯片的大数据存储与挖掘

It is one of the obstacles facing the big data that how massive heterogeneous data are stored and managed in the long term so as to realize efficient mining of data in different levels. DNA computing has the advantages of high parallelism and huge storage capacity, which provides an effective means to solve the problem. In this project, we will focus on exploring the new algorithms of DNA coding and the construction of molecular beacon micro-fluidic chip, and on this basis, study the storage and mining model of big data. Firstly, we will design the new DNA encoding algorithm by means of improving the constraint condition of DNA encoding, encode features of the data, and link linear DNA fragments using of DNA ligases, and build a DNA "hard disk" with lower volume, higher density, more stability that is suitable for the massive storing of information in the big data. Secondly, using the specific "hairpin" structure of molecular beacon, design the molecular beacon probe connected with fluorophore and quencher, that are complementary to the feature of the data to be extracted, and based on high parallelism of DNA computing, and combine with molecular biology techniques for the construction of molecular beacon microfluidic chip to realize feature extraction of the big data, propose the mining model of big data. Finally, the validity and efficiency of the model is verified via the biochemical experiments. The results of the project are sure to broaden the applications of DNA computing, and provide a potential means for the storage and mining of big data.

如何长久存储和管理海量异构数据并实现不同层次的高效挖掘，是大数据面临的难点之一。DNA计算具有巨大并行性和海量存储能力等优点，这将为该问题的解决提供有效途径。本课题将重点探索DNA编码新算法和分子信标微流控芯片制备，并以此为基础深入研究大数据的存储与挖掘模型。首先通过改进DNA编码约束条件设计DNA编码新算法，对数据特征进行编码，使用连接酶链接成线性DNA片段，结合DNA计算的海量存储能力构建具有体积小、密度大、稳定性强且适用于大数据存储的DNA"硬盘"。其次利用分子信标特有的“发夹”结构，将待提取数据特征的补链设计成连接有荧光基团的分子信标探针，依托DNA计算的巨大并行性，结合分子生物学技术制备分子信标微流控芯片以实现大数据特征的提取，构建大数据挖掘模型。最后通过生化试验操作，验证模型的有效性。本课题的研究不仅能拓宽DNA计算的应用范围和领域，也为大数据的存储与挖掘提供一种新途径。

课题组针对分子信标特有的“发夹”结构，阅读了大量相关文献资料；收集挖掘分子信标的信息表达方式，研究了分子信标对信息表述、信息存储及荧光标记等方面的优势并进行挖掘。进而课题组整理和研究分子信标在信息表达上的各种模型，开展用分子信标进行数据存储的研究。课题组利用分子信标的特点，设计分子信标自组装瓦片，构建分子信标自组装模型，扩展分子信标在数据存储的同时进行计算。接着利用分子信标可荧光标记的特点，结合DNA折纸术和杂交链式反应这两种新的自组装方法的进行计算模型的设计，给出其在组合优化和逻辑计算中的应用，为分子信标微流芯片在组合优化中的应用打下基础。研究后期，课题组结合e-learning中的教育资源个性化推荐，研究分子信标微流控芯片在教育资源个性化推荐中进行数据存储于挖掘的应用；总结研究成果，构建分子信标的存储和计算模型，提出分子信标在e-learning的教育资源个性化推荐中的应用设计策略；并发表了相关研究成果。